1. Field of the Invention
The present disclosure relates to an auto-focus device and method, especially to a passive auto-focus device and method.
2. Description of Related Art
The focus technique for a general image capture device includes active auto-focus technique and passive auto-focus technique. The active auto-focus technique is operable to actively send out a signal and receive the response thereof to determine the distance between an image capture device and a to-be-captured object while the passive auto-focus technique is operable to determine the optimal focus for the to-be-captured objet by analyzing the definition of the temporarily captured image of the said object.
Regarding the passive auto-focus technique, hill climbing algorithm is widely adopted in this kind of technique. The hill climbing algorithm, as implied by the name, finds out the peak of a definition variation curve in connection with a focus-searching area in the direction of ascending definition with a focusing mechanism, and takes the position of the focusing mechanism in association with the peak as the position for the optimal focus. The said focus-searching area is usually the central area of a to-be-captured scene, and the range thereof is usually constant. More specifically, if the range of the focus-searching area is too broad, there is a high possibility of including scenes in different depths of field; in this case, the hill climbing algorithm may choose the definition peak in connection with an indifferent scene and lead to an erroneous focus. On the contrary, if the range is overly narrow, there is a high possibility of missing the target scene; in this case, the hill climbing algorithm can only find out the definition peak in connected with an indifferent scene within the narrow range and miss the interested scene outside the range. In practice, a regular focus-searching range could be appropriate for certain scenes, but too broad or too narrow for others. As for other passive auto-focus techniques using different algorithms, they might face the similar problems, the problem of slow focusing speed, or the problem of noise influence.
People who are interested in the prior art may refer to the following document: Fang-Hsuan Cheng, Hsin-Wei Mao, “Research On Fast Image Based Auto Focus Technique”, Journal of Information Technology and Applications, Vol. 3, No. 1, pp. 67-76, 2008.